Visual-Approaches-for-Clustering-Tendency-Assessment-of-Graph-Data
This work presents two novel methods to cluster generic graph networks into communities using the VAT family of algorithms without any prior knowledge about the community structure. The two proposed methods recover the community structure with a high level of accuracy. The first of these methods is based on spectral graph theory. It finds the maximal eigengap corresponding to the eigenvalues of the graph connectivity matrix which is in turn used to compute a set of dissimilarity values. The second method introduces a pairwise dissimilarity score based on the number of common neighbors between two nodes of the graph.
Kindly use MATLAB software for executing the program. Execute the "main_prg.m" file.